Smartphone supported Activity Level Estimation

Jump to: navigation, search

Sensor-based decisions in eHealth

by Serhat Sama
Supervisor(s) Josef Noll, Zahid Iqbal, Dafferianto Trinugroho
Due date 2014/02/01
Status Finished
Problem description: In future Internet of Things (IoT) based systems a huge amount of data will be available, not only from these sensor systems, but also from the machine readable Web. The objective of this thesis is to establish a sensor platform, which will contribute to decision making. Such decision making might be part of business modelling, or simply a support system for users in context-aware situations. The selected scenario is based on eHealth, with a home-supporting system.
Methods and Tools: The tools and methods in this thesis are based on
  • A set of scenario, describing the challenges
  • A list of requirements being extracted from the scenarios
  • A description and evaluation of technologies and tools being candidates for solutions
  • A functional architecture/description of the envisaged system
  • An implementation of the core concepts
  • A demonstration of the solution
  • An evaluation of the solution, including a critical review of the descisions taken earlier
  • Conclusions
  • References
Time schedule The envisaged time schedule (for a long thesis/60 ECTS) is:
T0 0 starting month, T0+m denotes the month where the contribution to a certain chapter shalle be finalized
T0+2 months: create an initial page describing the scenario
T0+3: Provide a list of technologies which you think are necessary for the thesis
T0+4: Establish the table of content (TOC) of the envisaged thesis. Each section shall contain 3-10 keywords describing the content of that section
T0+7: Provide a draft of section 2 (scenario) and 3 (technologies)
T0+10: Establish a draft on what to implement/architecture
T0+11: Set-up an implementation, testing and evaluation plan
T0+15: Evaluate your solution based on a set of parameters, keep in mind there is no such thing as a free lunch
T0+17: Deliver the thesis
Pre-Knowledge This thesis includes a reasonable amount of programming. The envisaged thesis is based on radio communications, thus expects the user to have followed at least two radio-related courses
Approved Pending by
Keywords Context, Mobile, Sensor, eHealth, Shepherd, Telenor Objects, ANT+, Bluetooth

this page was created by Special:FormEdit/Thesis, and can be edited by Special:FormEdit/Thesis/Smartphone supported Activity Level Estimation

The thesis was delivered in February 2014.

Download thesis here


This thesis deals with the prototypical implementation of activity zone monitoring using the mobile phone. It uses the mobile phone sensors especially the accelerometer sensor to establish four types of motions and through a corresponding analysis with heart rate monitoring equipment, and then establishes the intensity of the activity. The goal behind the activity zone estimator as being suggested in this thesis is to provide notions of an activity of a certain intensity by only using the mobile phone without using external sensors. The implementation is based on first an analysis of existing technologies both when it comes to programming and when it comes to applications being available for mobile phones, and our implementation analysis then points out that the accelerometer is well tailored to establish an activity zone. However, the challenges might still occur with respect to, for example, the position of the smartphone on user’s body, elevation of the ground where user performs the activity, and battery life time.

Alternative Scenario

We are using an extended scenario from Dave, addressing the eHealth at home. We'd like to focus on positive monitoring, e.g. finding out if someone has not left the bed (sleeping room) or might have fallen in the bathroom. We will use sensor information for tracking people, typically in form of the position API from Android and some extra sensor information, e.g. WLAN field strength or infrared surveillance.

These data are collected and made available through the Telenor Objects (machine to machine - M2M) platform Shepherd. On top of the platform we will build an application, which uses decision making tools to inform other people.

Typical run-through with challenges

  • sensors collecting data, (e.g. mobile phone position, Bluetooth 4.0 sensors)
  • data collection into Telenor Objects (Master Yen)
  • build an application platform using Telenor Objects data
  • building the decision engine for "information"
  • display results as an Android app

Collaboration with external partners

EU nSHIELD (and pSHIELD) project, where we have deployed a sensor platform consisting of:

  • see pSHIELD-JBV-demo for detailed description of components
  • micro platform has: accelerometer, temperature, light + GPS position + time stamp
  • iPhone has: accelerometer (not used), position

Academic challenge:

  • idea about the security of a system, including an analysis of threats (attacks), mechanisms to protect (encryption) and system view (integration of components)

Telenor Objects

  • Telenor Objects has a focus on health care and tracking of people
  • they are provider of the Shepherd platform
  • talk to Juan Carlos Lopez Calvet for examples
  • existing sensor systems -> deliver data to Telenor Objects -> inform people
  • Challenge: "trust-based privacy"

UiA and

  • Dave Dafferianto.Trinugroho has developed the scenario, described in Media:Dave-eHealth-Scenario.pdf
  • he has also created the first .owl (ontology description for the home scenario)
  • check InnoMed for scenarios and other partners being involved
    • Invite yourself to seminars (17. january 2012, ...)


sensor comparison - various sensors to be integrated

  • Bluetooth 4.0 sensors (Mobile Phone, MacMini,...)
  • TiniDB and esper (complex event processing) -> algorithms to establish
  • Sensors in home-care systems (what happens in an environment)

Topics & Discussions



  • sensors collecting data, (e.g. mobile phone position, Bluetooth 4.0 sensors)
    • review/search for Bluetooth 4.0 sensors (related to eHealth?)
  • data collection into Telenor Objects (Master Yen)
    • Android tracker for Telenor Objects (app)
  • build an application platform using Telenor Objects data
    • access to the Telenor Objects platform (subscribe)
    • install a copy of results on our server (Zahid.Iqbal and Sarfraz.Alam)
  • building the decision engine for "information"
    • data analysis and modelling
    • semantic technologies: .owl, java api, reasoning (Dave)

  • display results as an Android app
    • Priority handling, "emergency" -> flash information, "look after" -> reminder, "normal" -> status
    • compare to Bipper app (


  • access control for data depending on context, content (value) and trust
    • semantic technologies
  • "virtual sensor" (combination of several sensor data) (++)


Title page, abstract, ...

1. Introduction, containing: short intro into the area, what is happening (mobile phone, elderly generation, technology support, ...)
1.1 Motivation, containing: what triggered me to write about what I'm writing about
1.2 Methods, containing: which methods are you using, how do you apply them
2. Scenario, optional chapter for explaining some use cases
2.1 user scenario,- eHealth home support
2.2 Requirements/Technological challenges - described them under Topics
3. State-of-the art/Analysis of technology, structure your content after hardware/SW (or other domains). Describe which technologies might be used to answer the challenges, and how they can answer the challenges
3.1 Sensor description, typically Bluetooth 4.0 and Mobile Phone
3.2 Android programming environment: App, location information
3.3 Telenor Objects platform - Shepherd
4. Implementation
4.1 Architecture, functionality
5. Evaluation
6. Conclusions

Tips & References